Document-Level Supervision for Multi-Aspect Sentiment Analysis Without Fine-grained Labels

TL;DR

This paper introduces a VAE-based method for multi-aspect sentiment analysis that uses only document-level supervision, eliminating the need for detailed aspect and sentiment annotations, and achieves superior results on benchmark datasets.

Contribution

It presents a novel approach that leverages document-level labels for multi-aspect sentiment analysis without requiring fine-grained annotations.

Findings

Significantly outperforms state-of-the-art baselines on benchmark datasets.

Effectively detects multiple aspects within documents.

Utilizes only document-level supervision for training.

Abstract

Aspect-based sentiment analysis (ABSA) is a widely studied topic, most often trained through supervision from human annotations of opinionated texts. These fine-grained annotations include identifying aspects towards which a user expresses their sentiment, and their associated polarities (aspect-based sentiments). Such fine-grained annotations can be expensive and often infeasible to obtain in real-world settings. There is, however, an abundance of scenarios where user-generated text contains an overall sentiment, such as a rating of 1-5 in user reviews or user-generated feedback, which may be leveraged for this task. In this paper, we propose a VAE-based topic modeling approach that performs ABSA using document-level supervision and without requiring fine-grained labels for either aspects or sentiments. Our approach allows for the detection of multiple aspects in a document, thereby allowing for the possibility of reasoning about how sentiment expressed through multiple aspects comes together to form an observable overall document-level sentiment. We demonstrate results on two benchmark datasets from two different domains, significantly outperforming a state-of-the-art baseline.